OpenAI reports six AI misbehavior incidents, including fake citations and oversight escapes

The company has published six previously undisclosed incidents involving its AI models and said future reporting will cover ...